Analisis Kepuasan Pelanggan Melalui Sentimen Ulasan Menggunakan Algoritma Naive Bayes

Abstract

The advancement of digital technology has led to a large amount of data being generated by active internet users. In this study, the focus is on analysing customer sentiment towards MSMEs selling on the Tokopedia platform. The analysis is conducted using the naïve bayes algorithm method and Net Promoter Score (NPS) to measure customer satisfaction. The findings reveal that 75.69% of customer sentiment is positive, 11.08% is neutral, and 13.23% is negative. The Multinomial Naive Bayes model utilized achieves an accuracy rate of 80%, with precision, recall, and F1-Score all reaching 80%. Moreover, the calculated NPS value is 62.46%, indicating that a majority of customers are satisfied with the MSMEs on Tokopedia